IRJET- Embedding Randomness into Symmetric Key Encryption using Genetic Algorithm

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 03 | Mar 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 5121
Embedding Randomness into Symmetric Key Encryption using Genetic
Algorithm
Globy Yohannan1, Prof. Soumya J.W2
1PG Scholar, Dept. of ECE, Believers Church Caarmel Engineering College, Perunnad, Kerala, India,
2Assoc. Prof, Dept. of ECE, Believers Church Caarmel Engineering College, Perunnad, Kerala, India
----------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Data security is equally important as the storing
and communication of data. Shared key encryption is a
method that provides security by using pre-shared key. It
generates the same cipher text for the same set of plain text
and key. Security can be improved by imposing some sort of
randomness in the key. This can be achieved by using genetic
algorithm. Using this algorithm different cipher text can be
generated for same set of plain text and key. Because of the
randomness added to cipher text generation, standard attack
methods such as brute force, chosen-plaintext attacks etc. are
made more difficult. The effectiveness of the algorithm is
tested on plain text, and speech and results are satisfactory.
Key Words: Shared key encryption, Randomness,
Embedding, Extracting, Genetic algorithm, Data security.
1. INTRODUCTION
Nowadays, communication technologies have become the
medium to exchange the data in the form of text, image and
audio. So, data security is as important as that of the
collection of information. To avoid eavesdropping,
encryption methods are needed. The overhead added to
communication and storing of sensitive data, due to
cryptographic methods, is a necessary side-effect for the
protection of valuable information. Cryptography schemes
ensure data security and avoid modifications by the
unauthorized users. Mainly there are secret key encryption
and public key encryption. Shared key encryption provides
data security by using a pre-shared key. The pre-shared key
is exchanged through a secure medium. In shared key
encryption, same key is used for both encryption and
decryption. Public key encryption method ensures data
security by using two keys (public key and private key).
Public key can be accessed by anyone, but private key can
only be accessed by the owner. Public key encryption
method provides more security than shared key encryption.
To provide the same level of data security as that of the
shared key encryption method, Public key encryption
method is computationally cost and complex. So, it is
necessary to generate an unbreakable algorithm, which is
suitable for less resource consumption devices (that is, low
cost in terms of computation and communicationoverhead).
Same cipher texts are generated for same set of plain text
and key using shared key encryption. Once the shared key is
known to the intruder the entire security will compromise.
To increase the security using shared key, different cipher
texts are generated for same set of plain text and key. And
security can be improved by imposing some sort of
randomness in the key. Randomness can be provided by
using genetic algorithm operators. It generates unique and
minimum correlated (i.e., maximum difference) cipher text.
Randomness imposed by genetic algorithm increases
security but it introduces another problem, theintermediate
key generated by the sender side is not known to the
receiver side. So, we need to embed the intermediate key
along with the cipher text (Note thatthesecurityheremeans
is the analytical security not the security of the symmetric
key encryption). Because of the randomness it is made
difficult to apply the standard attack methods such as
dictionary method, brute force, known key method etc. It is
made impossible to find which function is used for
encryption because of the randomness. This method is
suitable for IoT, mobile andlesscomplexapplications.Thisis
helpful in improving security in cryptographic algorithms in
embedded systems without increasing computational
complexity manifolds.
The security level of any symmetric key algorithmisdirectly
proportional to the execution time [2]. This canbeimproved
by increasing the number of iterations(stages)ofsharedkey
encryption. Genetic algorithm finds the fittest solution for
the given problem using genetic operators. In this paper, a
continuously changing random key is generated with the
help of G.A. So standard attack methods can’t retrieve
information from the encrypted data or cipher text. Unless
the secret key is known to hacker, it is impossible to decrypt
the text because different ciphertextaregeneratingforsame
set of plain text and key each time an iteration of the
algorithm is carried out. Both the key and data are secured
using this method. This is similar to embedding information
into images using steganography.
2. GENETIC ALGORITHM
Genetic algorithms are a class of optimization algorithm [4],
which finds the suitable fittest solution for a given problem
from a population. It is based on natural selection,
inheritance,mutation,andcrossover.Mutationandcrossover
are the genetic algorithm operators in G.A. The genetic
algorithm belongs to the family of evolutionary algorithms,
along with genetic programming, evolution strategies, and
evolutionaryprogramming[3].Geneticalgorithmfindsalotof
application in cryptanalysis field, mathematics, engineering
etc. Genetic algorithm considers an optimization problem as
the environment, where feasiblesolutionsaretheindividuals
living in that environment [1]. Population contains a several

solutions or individuals or chromosomes. Chromosome
means the possiblesolutiontoparticularproblem.Eachvalue
of chromosome is called the genes.
Genetic algorithm generates the initial populationrandomly.
Once the initial population is generated, it enters the loop to
generatenew population. New populationisgeneratedatthe
end of the loop. Dependingupontherequirementthenumber
of loops can be increased to find the fittest solution.
Regeneration process occurs with the help of genetic
algorithm operators. Operators are selected based on the
application. In this paper, single point crossover and
mutation are used.
After the generation of the initial population, fitness of each
individual is calculated. Based on the result of the fitness
calculation the best fittest solutions are selected from the
population. These solutions are taken as the parents for the
next generation. Newoffspringaredevelopedwiththehelpof
reproduction operators. It selects two or more parents and
after certain recombination new individuals are generated
with crossover operator. Modification of individuals is done
with the help of mutation operator. It helps to maintain
diversity. These offspring along with the parents are used as
the next population. Parents are along with the offspring as
the next population because if we are using offspring alone
for the next stage, we can’t guarantee that the new offspring
will be better than the previous.Sometimesthenewoffspring
may be worse than the previous offspring. Using the parents
along with the offspring wecan assurethat the newoffspring
will be as better as the previous one.
Genetic algorithm is used instead of pseudorandom number
generation because, in the case of pseudorandom number
generator, it generates number that seems to be random but
actually not. G.A is an optimization class and it generates
number that is actually random using genetic operators. The
fitness function in our implementation of genetic algorithm
prefers (selects) chromosomes or individuals which have
minimum correlation and maximum difference with other
individuals of the same generation. This is what we need in
this paper.
3. PROPOSED SCHEME
Genetic algorithm along with embedding is used in the
proposed scheme to improve the security of the sensitive
information. Random keys are generated with the help of
genetic algorithm. It generates unique cipher for same set of
plain textand key each time the algorithm is carried out. And
the random keys are made continuously changing. Our
implementation supports multiplesimultaneousinstancesof
the algorithm. This algorithm is suitable for text, image and
speech file.
In the case of text file the eight bit is vacant (since for inputs
A-Z, a-z, 0-9 and common symbols the 8th bit remains
vacant). But for the case of special characters or images or
text the 8th bit is not vacant. And there is a chance of bit
overflow this can be overcome by using subtraction
operation. The encryption process occurs as three stages.
Secret key is first shared between sender and receiver
through a secure medium. In the first stage:
1. The inputs to this stage are plain text and random
key.
2. Random keys are developed with the help of genetic
algorithm.
3. Plain text is divided into nibbles (means as 4 bits).
Left nibble of plain text is X-ored with the left nibble
of random key.
4. Then the right nibble of the key is subtracted from
the right nibble of the above result.
5. Because of subtraction there is a chance of getting
signed values. To convert the signed values into
unsigned values take 2’s compliment.
6. Then the right nibble of the result of 2’s compliment
is X-ored with the right nibble of the random key.
7. The resultant is given as the input for next stage of
encryption.
Fig-1: Block diagram of Encryption stage
(We are taking the input as 8 bits). The reason for using Xor
instead of OR and AND operation is that, using OR and AND
operation encryption process is possible but during the
decryption stage we can’t get the exact plain text back. Using
Xor operation the exact plaintext can be retrieved duringthe
decryptionprocess.XorOperationisalsoknownasreversible
process. In the second stage of encryption:
1. The input to this stage is the cipher text generated in
the first stage (cipher text 1).
2. The bits in the primeFibonaccipositionsofthecipher
text are complemented (that is, Xor cipher text with
the hexadecimal value 3A).
3. Then the entire cipher text is reversed.
4. After the reverse operation once again, the bits inthe
prime Fibonacci positions of the above result is
complemented.
5. Then the result is divided into nibbles. The leftnibble
is storedas one dimensional array and Xor operation
is performing on the right nibble with reverse of the
left nibble (That is, the 1st bit is Xored with the 8th

bit and store the result in the 8th bit, The 2nd bit is
Xored with the 7th bit and stored it in the 7th bit and
continue the process till 4th bit is Xored with the 5th
bit and store the result in the 5th bit).
6. The result obtained is given as the input to the third
stage of encryption process.
The reason forusing the Fibonacci primeposition is that, itis
widely seen serieslikethefundamentalsignalssinewaveand
cosine wave. It is possible to predict the growth using this
series. In the Stage three encryption process:
1. The inputs to this stage are cipher text generated
from stage 2 encryption (cipher text 2), shared key
and random key.
2. In this stage the simultaneously generating random
keys are embedding in the cipher text based on the
shared key sequence.
3. Then the cipher text is converted into ASCII value.
Fig-2: Block diagram of Decryption stage
Embedding stage occurs as follows: random keys are
embedded into the cipher text based on the shared key
sequence. If the shared key shows 0 as value, embed a bit
from cipher text. If the sharedkey valueis1,embedabitfrom
random key. Continue the processtillthereremainsnobits.If
the shared key value is exhausted first, continue to scanfrom
the start. For the case of image, reshape the 3-D array (two
dimensions plus color data) into 1-D array and embed any
dimension into 1-D array which is needed in the decryption
process. In the decryption stage the plain text is retrieved
from the encrypted data. Decryption occurs as three stages.
In the decryption stage 1:
1. The inputs to this stage are cipher text 3 and shared
key.
2. In this level random key and cipher text 2 are
extracted from the cipher text 3.
Extraction is the reverse process of embedding. In the
extraction process cipher text and random keys are
extracting based on the sequence of the shared key. Assign 0
as cipher text and 1 as random key or vice versa. And start to
scan the shared key, if the bit of shared key is 0, store the bit
from the cipher text 3 as cipher text 2. If the bit of shared key
is 1, store the bit from the cipher text 3 as random key.
Continue the process till all the bits of the cipher text 3 gets
exhausted. If the shared key is exhausted first, then continue
to scan from the first position. In the decryption stage 2:
1. The input to this stage is cipher text 2.
2. The input is divided into nibbles. The left nibble is
stored as one dimensional array and Xor operation
is performing on the right nibble with reverse of the
left nibble (That is, the 1st bit is Xored with the 8th
bit and store the result in the 8th bit, The 2nd bit is
Xored with the 7th bit and stored it in the 7thbitand
continue the process till 4th bit is Xored withthe5th
bit and store the result in the 5th bit).
3. The bits in the prime Fibonacci positions of the
cipher text are complemented (that is, Xor cipher
text with the hexadecimal value 3A).
4. Then the entire cipher text is reversed.
5. After the reverse operation once againthebitsinthe
prime Fibonacci positions of the above result is
complemented.
Cipher text 1 is regenerated after this stage, which is givenas
the input to next stage. In the decryption stage 3:
1. The inputs to this stage are ciphertext1andrandom
key.
2. The cipher text 1 is divided into nibbles. And Xor
operation is performed on the right nibble of cipher
text with right nibble of the random key.
3. Then the 2’s compliment of the result is taken to
convert the unsigned integertosignedintegervalue.
4. Then the addition operation is performed between
right nibble of the cipher text and the right nibble of
the random key.
5. Xor operation is performed between left nibble of
the cipher text and the left nibble of the key.
The plain text is generated after this stage.
3. RESULTS
The proposed algorithm is tested on text, image and speech
file. Results obtained are quite satisfactory. Below shows the
results obtained for a text file using proposed algorithm.
Algorithm is tested using text file as sentence and as
paragraph. Plain text and shared key given are:
Plain text: Chrysanthemums were first cultivated in China as
a flowering herb as far back as the 15thcentury BC. Over 500
cultivars had been recorded by 1630. The plant is renowned
as one of the Four Gentlemen in Chinese and East Asian art.
The plant is particularly significant during the Double Ninth
Festival.
Shared key: @Farenheit320.

Random keys are kept changing continuously. The following
shows the generated randomkeysandtheencrypteddatafor
the given input:
Random key: 0-#;[Cc0 and - nxGmh.c
Cipher text:
9 Œ1<K˜¬& fR' lŽæâH°#Õ ÃÀ€ æbRÁ4AÆ4V¥$V ‘" !
Vïã4dV†áÑ1 Ò F¤óu ¢R€q FT… äÖ2‚ ÐÁ & ¶¿äÖ2"`
ÓT¥oä5 á# Ä‘Ä%²Ïçö QÅå Â ¶T£%UbÒÁ1 F ¡ãÄ
æâÀ 6÷#¥TåÅ æÕ2Âïä$d åƒ¡Ð„AÆ”±äÄ ‡ñr€ AÆ ±äõ
… ƒò äEF_ãDó Â€r „F†TÕT€åP& øA?Ûp 4Ë@ ç¿±ß¡ •
W< À5¾¹1 á ³_ÖïåÑ …åâ …ŸÀæÃ• CÁ’³à Ô¡&“¶“Ïß" ³P
vã ‚° £ ¤•Æ”Åv€••“áFôŸÆÃQáïQâ¯c ƒ‘7C ‚Ÿ åæÖ
5. CONCLUSION
The data security using symmetric key encryption can be
improved by imposing some sort of randomness with the
help of genetic algorithm. Finding which function is used for
encryption using standard attack methodsaremadedifficult
using this method. The proposed algorithm is tested for text
and speech files and results obtained are quite satisfactory.
REFERENCES
[1] Ajay Kr. Phogat and Archana Das, “A Symmetric
Cryptography Based on Extended Genetic Algorithm,”
IJCTER, Volume 2 Issue 4, April 2016, pp. 541-547.
[2] Subhajit Das, Satyendra Nath Mandal and Nabin
Ghoshal, “Diffusion and Encryption of Digital Image
Using Genetic Algorithm,” FICTA, Volume 1, 2014.
[3] Suvajit Dutta, Tanumay Das and Sharad Jash, “A
Cryptography AlgorithmUsingthe OperationsofGenetic
Algorithm & Pseudo Random Sequence Generating
Functions,” IJACST, Volume 3, No. 5, May 2014.
[4] Aarti Soni and Suyash Agrawal, ”Using Genetic
Algorithm For Symmetric Key Generation in Image
Encryption,” IJARCET, Volume 1 Issue 10, December
2012.
[5] Sindhuja K and Pramela Devi S, “A Symmetric Key
Encryption Technique Using Genetic Algorithm,”IJCSIT,
Volume 5, 2014.
[6] Rasul Enayatifar and Abdul Hanan Abdullah, “Image
Security Via Genetic Algorithm,” IPCSIT, Volume 14,
2011.

IRJET- Embedding Randomness into Symmetric Key Encryption using Genetic Algorithm

Recommended

More Related Content

What's hot (20)

Similar to IRJET- Embedding Randomness into Symmetric Key Encryption using Genetic Algorithm (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- Embedding Randomness into Symmetric Key Encryption using Genetic Algorithm